The Quest for a Better Setup: Part One

Last summer, I spent most of my time working on an analysis of changes within the Grim Dell during periods of rainfall.  In the very last week, I embarked upon a microcosm project. I followed the setup of a previous lab member, freezed my samples and analyzed them during the 2013-2014 school year.  My results were generally inconclusive, and during the school year I had a lot of time to consider the ways that the project could be improved.

I felt that the established setup didn’t allow for homogeneous replicates and that the process of filtering viral concentrates could be optimized.

It’s one thing to see areas that could be improved and another to actually improve them.

Homogenization was the first problem we tackled,  “we” being the entire lab group.  One of my favorite things about the way the Williamson lab functions is that we all feel comfortable coming to each other for suggestions, advice, and general perspectives on our work.  I came in to the lab at the beginning of the summer and told my labmates my problem.  I felt like the samples were not being filtered well before being added to microcosms, and things were settling to the bottom of the carboy during transfers.  This needed to change.  One of my thoughts was use of a 100 micron filter to transfer over liquids, and a labmate suggested use of a bubbler (similar to the type you find in an aquarium) to maintain movement in the larger carboy.  The group discussion really opened up avenues of thought for me, and we eventually decided that the 100 micron filter would work for pre-filtration, and use of a peristaltic pump would work to keep water moving from carboy to carboy.

Once a solution was derived, we had to buy the supplies to make it happen.



  1. Denise Lee says:

    Great job on coming up with a solution to a problem! Has the 100 micron filter fixed the issue and have you had better data with the new change?

  2. It appears to have made a significant difference, and I feel that the data collected this time around is far more reliable as a result of the additional step.